Unsupervised Reservoir Computing for Solving Ordinary Differential Equations

• Published in: arXiv.org
• Main Authors: Mattheakis, Marios, Hayden, Joy, Protopapas, Pavlos
• Format: Paper
• Language: English
• Published: Ithaca Cornell University Library, arXiv.org 25.08.2021
• Subjects: Back propagation; Back propagation networks; Bayesian analysis; Differential equations; Initial conditions; Linear equations; Neural networks; Nonlinear systems; Optimization; Ordinary differential equations; Parameters; Recurrent neural networks; Robustness (mathematics); Solvers

There is a wave of interest in using unsupervised neural networks for solving differential equations. The existing methods are based on feed-forward networks, {while} recurrent neural network differential equation solvers have not yet been reported. We introduce an unsupervised reservoir computing (...